Severe dyslipidemia is linked with higher intracellular proliferation of LD-R amastigotes.

(A) Serum lipid profile of 28 days LD-infected Balb/c mice (N=5). Equal number of sorted metacyclics promastigotes of two representative LD-S (LD-S1, LD-S2) and LD-R (LD-R1, LD- R2) strains were used to perform independent infection. (B) Number of attached LDs on KC surface 4Hrs p.i, data represents (N=3) for two LD-S and two LD-R strains. C (i) Representative Giemsa stain images of LD-S1 and LD-R1-infected KC at 4Hrs and 24Hrs p.i Black arrow represents LD amastigotes. N denotes host nucleus. Scale bar 20μM. Giemsa Images are represented in grey scale to clearly represent LD nucleus (black arrow). C (ii) Intracellular amastigote count for 4Hrs and 24Hrs p.i. Each dot represent count from 100 infected-KC (N=6). D (i) LD-infected-KC imaged to visualize the PV at 4Hrs and 24Hrs p.i. PV marked by ATP6V0D2 (magenta), KC and LD nucleus (blue), Scale bar 5μm. Yellow arrow represents LD amastigotes. In the merged image, white dotted line marks KC’s periphery, and host nucleus is represented with N. D (ii) PV counts measured from 30 infected KCs. D (iii) 3D morphometric representation for better visualization of PV in 24Hrs post infected KC marked with white dotted box as in D (i). ATP6V0D2 (magenta), nucleus (cyan) with last panel showing PV-grooves marked in red solid line with white arrow representing LD amastigotes. *** signifies p value < 0.0001, * signifies p value < 0.05, n.s non-significant.

LD-R-infection results in a higher organ parasite load in hypercholesterolemic Apoe-/-mice. A

(i) Representative Giemsa images of LD-infected-KCs cultured in high and low lipid media. Black arrow represents LD amastigotes; N represents host nucleus. Scale bar 20μM. A (ii) Graph representing number of intracellular LD amastigotes in KCs with LD infection performed either in high, low or lipid free media (N=6). B (i) Giemsa images of LD-infected-KCs isolated from wild type BL/6 and Apoe-/-mice. Scale bar 20μm, N representing host nucleus. Giemsa Images are represented in grey scale to clearly represent LD nucleus (black arrow). B (ii) Amastigote count from 100 LD-infected-KCs (N=6) as in B(i). Data are presented as mean ± SEM. C (i) Schematic representation of in-vivo LD-infection in BL/6 and Apoe-/- mice performed with EGFP-LD-S or EGFP-LD-R C (ii) Flow cytometry representing GFP-positive splenocytes isolated from 11 days EGFP-LD infected BL/6 and Apoe-/- mice. The black box indicates % of GFP-positive splenocytes (LD-infected). C (iii) Graph representing data from six infected mice as in C (ii) are presented. *** signifies p value < 0.0001, ** signifies p value <0.001, * signifies p value < 0.05, n.s non-significant.

Endocytosed-LDL is the primary lipid source for intracellular LD-R proliferation.

A (i) Live cell images of LDL-endocytosis in 24Hrs post infected KCs with LDL (red), LD (green), and the nucleus (blue). Images were acquired directly from 96 well glass bottom plate and bright field was merged with reduced brightness and increased contrast identically for all experimental conditions. Scale bar 40μm, Yellow arrow represents intracellular LD. A (ii) Graph representing Fluorescence plate reader-based quantification of A (i). B (i) Co-localization of LDL with the LD- amastigotes in infected KCs 24Hrs p.i. Merged image showing LDL (magenta), LD (green), nucleus (blue), White dotted line in the merged image demarcates infected-KC for co-localization analysis. Scale bar 10μm, yellow arrow represents LD nucleus and N represents host nucleus B (ii) Intensity plot representing co-localized signals from B (i), marked with grey transparent box. C (i) GC-MS lipid profile of isolated PVs from LD-infected-KCs. Retention time 27.73 corresponding to the cholesterol and 28.56 corresponding to fatty acids are represented with * for LD-S-PV and with black arrowheads for LD-R-PV. C (ii) Relative abundance of Cholesterol and Fatty acids of C (i). D (i) Representative Confocal Raman spectroscopy image of uninfected or LD-infected-KCs 24Hrs p.i, illustrating distinct lipid-related signal distribution patterns marked with pseudo-colours. Yellow dotted circle in the left most panel demarcates the host cell nucleus while white arrow marking intracellular LD nucleus. In merged panel white dotted line demarcates cells periphery. The right most panel shows dot plot representation of lipid distribution, where each colour corresponds to different lipid-related peaks with respect to blue (B) indicating the nucleus. D (ii) Comparative Raman spectra from LD-S and LD-R-infected-KCs. Lipid-related peaks are demarcated with shaded box, representing 1. 540-560 cm-1 (Cholesterol), 2. 1080-1090 cm-1 (phospholipids), 3. 1270-1280 cm-1 (triglycerides), 4. 1300-1340 cm-1 (Amide-III bond), 5. 1440-1453 cm-1 (Fatty acids and Triglycerides), and 6. 1650-1660 cm-1 (Amide-I bond). *** signifies p value < 0.0001, ** signifies p value <0.001, * signifies p value < 0.05, n.s non-significant.

Increased LDL uptake in LD-R-Infected-MFs through receptor independent fluid phase endocytosis.

(A) Immunofluorescence images of LDLr expression in LD-infected-KCs. Scale bar 40μm. Enlarged view of A showing infected-KC with low LDLr expression (yellow arrow) compared to neighboring uninfected KC (white arrow) in the merged image. B (i) Cofilin and phosphorylated-Cofilin expression by western blot in LD-infected-KCs. B (ii) Graphical representation of comparative expression of Cofilin and phosphorylated-Cofilin with respect to β- actin for B (i). C (i) SIM images showing the Cofilin distribution in LD-infected-KCs. Yellow arrow represents intracellular LD, white dotted line in the merged image demarcates cell boundary. Scale bar 5μm. C (ii) Colour spectrum representing Cofilin distribution pattern of C (i) with green representing maximum and red representing minimum distribution. D (i) Confocal images representing Filamentous actin (F-actin) protrusions in LD-infected-KCs. Scale bar 20μm, with yellow arrow representing LD amastigotes. Individual experimental condition is represented in pseudo-colour, with uniform high contrast to clearly represent the F-actin protrusions. D (ii) Graph representing Cofilin and F-actin distribution in KCs infected with LD or left uninfected. (E) Still images from live cell video microscopy representing co-localization of LD-R with LDL in presence or absence of Latrunculin-A (LAT-A). Scale bar 60μm. Yellow arrow represents LD-infected-KCs with LDL(red). (F) Graphical representation (N=6) of amastigote count in 100 LD-R-infected-KCs treated with either LAT-A, or Cytochalasin (CYT-D) or left untreated. Data are presented as mean ± SEM. ** signifies p value <0.001, * signifies p value < 0.05, n.s non-significant.

Endocytosed-LDL fuses with LD-R-PV to provide Cholesterol to LD-R amastigotes.

(A) Confocal images of representing LDL-vesicles, with LAMP1 positive lysosomal vesicles in LD- infected-KCs 4Hrs p.i. Yellow arrow represents LD nucleus, white dotted box marks the region further cropped and zoomed to show fusion of lysosomal vesicles with LDL and LD-R. Scale bar 10μm. A (ii) Schematic representation showing trafficking of LDL-vesicle towards the LD-R-PV. B (i) SIM images representing convergence of LDL-vesicles with LD amastigotes and LAMP1 in 24Hrs with LD-infected-KCs. White dotted box marks the region further cropped and Zoomed to clearly show the convergence (of LD amastigotes with LDL and LAMP-1 (white vesicles shown by yellow arrow), Scale bar 5μm. B (ii) Intensity plot representing of co-localization signals for B (i). C Western blot analysis showing LAMP1 expression in LD-infected-KCs from early to late hours. D (i) Filipin staining of EGFP-LD-infected-KCs. Filipin (white) and LD (green), yellow arrow represents intracellular LD. Scale bar 10μm. White dotted line marked cells used to generate co-localization plot. D (ii) Co-localization signals of LD (EGFP and Filipin represented by the dot plot. Pearson co- localization coefficient value (r) represented for LD-S and LD-R-infected-KC. D (iii) Graph representing Pearson’s coefficient for EGFP-LD-infected-KCs (N=19). Statistical significance in the observed differences was determined through ANOVA utilizing GraphPad Prism software (version 9.0). *** signifies p value < 0.0001.

LD-R infection suppress NPC-1 to alter Cholesterol mobilization in infected MFs.

(A) 24Hrs LD-S and LD-R-infected-PECs were subjected to RNAseq analysis keeping uninfected- PECs as control. Heat map representing average expression of differentially expressed genes related to lipid metabolism between LD-S and LD-R-infected-PEC. Scale represents median- centered counts (TPM). Row Z score data represented here. Expression of RNAs (right margins) is presented as centered and ‘‘scaled’’ [mean normalized log2(TPM + 1)]. B (i) Expression of SREBP2 and B (ii) NPC-1 in LD-infected and uninfected KCs by Western blot with β-actin as control. (C) Intracellular localization of SREBP2 in response to LD-infection in KC 24Hrs p.i. White dotted line represents cell periphery with yellow arrow representing intracellular LD and N representing host nucleus. 3D image represented in the right most panel with one quadrant clipped to confirm nuclear translocation of SREBP2. Scale bar 5μm. (D) Immunofluorescence showing NPC-1 expression in LD-infected-KCs. White arrow representing NPC-1 expression in LD-R- infected-KCs and yellow arrow representing NPC-1 in adjacent uninfected-KCs in the merged panel. White dotted box marks regions further cropped and shown in the bottom panel with enlarged view (Zoomed). Scale bar 20μm. E (i) TIRF microscopic images of Laurdan stained LD- infected and uninfected KCs with two different spectra (488nm, 594nm), changes from red to green represents gel to fluid phase transition of the host plasma membrane. Colour distribution pattern corresponding to the TIRF images representing each experimental condition is presented in the right most panel. Scale bar 5μm. F (i) Expression of NPC-1 determined by immunofluorescence in PECs transfected with NPC-1 siRNA or scrambled (sc) siRNA. Scale bar 5μm. F (ii) Giemsa- stained images of LD-infected NPC-1 knockdown-PECs with scrambled siRNA control (Sc). Scale bar 20μm. Images are represented in grey scale with increased contrast to represent LD nucleus (black arrow). F (iii) Graphical representation of number of intracellular LD-amastigotes as in F (ii). Results show the counting of 100 infected-PEC from 6 independent replicate. * signifies p value < 0.05, n.s non-significant.

LD-R-infected-MFs selectively excludes ox-LDL to suppress inflammatory response in host.

(A) Scheme representing ex vivo co-culture experiment performed by isolating T-cells from LD-infected mice with in vitro LD-infected-KCs B (i) IFN-γ level in the supernatant of LD-infected-KCs co-cultured with the T-cells isolated from spleen of Balb/c mice infected with LD- S or LD-R 28 days p.i. Data represented as mean ± SD from three independent experiments. B (ii) IFN-γ level in the supernatant of splenocytes cultures from LD-infected-BL/6 and Apoe-/--mice stimulated with SLA (Soluble Leishmania antigen) specific for LD-S or LD-R. Data are presented as mean ± SD from six independent mice. C (i) ox-LDL incubated LD-infected-KCs was fixed and stained with Oil-red-O and Haematoxylin. LD-SHigh represents MOI (1:20) and LD-RLow represents MOI (1:5). Scale bar 10μm. Black arrow represents LD nucleus. C (ii) Graph representing % of ox- LDL positive MFs under different experimental conditions as represented in C (i). D (i) Confocal image showing MSR-1 in LD-infected-KCs. Yellow arrow represents LD-amastigotes while N represents host nucleus. Scale bar 20μm C (ii) Expression of MSR1 detected by Western blot of LD-infected-KC with β-actin as loading control. (D) Expression of ATP6V0D2 in LD-infected-KCs by Western blot with β-actin as loading control. E (i) Expression of MSR1 and neutral lipid droplet accumulation assayed by Confocal microscopy from the cryo-sectioned liver of 28 days LD- infected-mice. Representative image showing Lipid droplet marked in red and MSR1 in green, nucleus in blue. Scale bar 40μm. E (ii) Graphical representation of % positive Lipid droplets accumulation and MSR-1 expression under different experimental condition as in E (i). Data collected from N=6 mice. *** signifies p value < 0.0001, ** signifies p value <0.001, * signifies p value < 0.05.

Lipid droplets accumulation around LD amastigotes is inversely related with susceptibility towards Amphotericin-B.

(A) Ex-M image showing lipid droplet assimilation in LD- R-amastigotes. White dotted region further cropped and shown in enlarged panel (Zoomed). Nile red (magenta), LD (green), nucleus (cyan). White spots represent Lipid droplets assimilated in LD- R and Host nucleus is represented as N. Scale bar 5μm. (B) TEM images showing lipid droplets around the LD-R amastigotes. Black box further magnified and represented in right panel. White * representing LD-R amastigotes, yellow * showing lipid droplets. Scale bar 2μm. (C) Schematic representation representing experimental scheme used to determine EC50 against Amphotericin- B under different experimental conditions. D (i) and (ii) Determination of EC50 value against Amphotericin-B in low lipid and high lipid media by performing infection with of two independent LD-R strains. D (iii) Determination of EC50 value against Amphotericin-B in KC infected with LD- SAmp-B-R. Intracellular amastigotes were counted through Giemsa staining. (E) Confocal images showing lipid droplets accumulation around intracellular amastigotes with Aspirin (5 μM) or Amp-B (0.36 μM) or with a combination of Aspirin (5μM) and Amp-B (0.11μM )at 48Hrs p.i (F) Determination of amastigote load in KCs infected with two independent LD-R isolates in different experimental conditions either untreated or treated with Aspirin (5 μM) or Amp-B (0.36 μM) or with a combination of Aspirin (5μM) and Amp-B (0.11μM ) at 48Hrs p.i (G) Comparative Lipid profile of VL patients (Cured and Relapsed) in response to Amp-B treatment along with healthy individuals from endemic region of Bihar, India. *** signifies p value < 0.0001, ** signifies p value <0.001, * signifies p value < 0.05.

Scrutinized uptake of host lipids results in breakdown of Amphotericin-B responsiveness in clinical LD isolates with primary resistance to antimony.

(1-2) LD-R residing in membrane bound parasitophorous vacuoles (PVs), and (3-5) induce high fluid phase endocytosis of LDL by modulating actin depolymerization of host membrane through cofilin. (6-7) LDL containing vesicles fused with LD-R-PV which provides increased lipid access to sustain their aggressive intracellular proliferation. Vesicular cholesterol export protein, NPC-1 is downregulated in the fused membrane, preventing release of LDL derived cholesterol back to host cytoplasm. (8- 12) LDL derived cholesterol esters also recycled in the ER and form lipid droplets which ultimately accumulate in and around the LD-R-PV ultimately leading to the Amphotericin-B unresponsiveness. (13-15) LD-R control inflammatory response related to accumulation of oxidized lipids (Ox-LDL) by downregulating MSR-1 (16) High LDL influx also results in shutdown of de-novo cholesterol synthesis in LD-R infected host through restriction of nuclear translocation of SERBP2. Right panel, representing LD-S infected host with low LDL influx with high NPC-1 expression leading to compromised lipid droplet accumulation around its PV, thereby exhibiting high responsiveness against Amphotericin-B.

EC50 against SAG (Antimonial) for the LD strains used in this study

Lipid profile of VL patients

Kupffer cells (KC) as an in vitro model for LD-infection.

(A) Flow Cytometry representing % Metacyclics LD: (i) LD-S (AG83, DD8) and (ii) LD-R (BHU814, BHU575) based on FSC and SSC. Left, small box with low FSC and SSC gate for metacyclic parasites and right panel with high FSC and SSC gated procyclic parasites. (B) Flow cytometry of F4/80-PE positive KC from the non-parenchymal cell population isolate from murine liver. C Representative Giemsa-stained images LD-infected-PEC, 4 and 24Hrs p.i for (i) LD-S and (ii) LD-R strains. Giemsa Images are represented in grey scale to clearly represent LD nucleus (black arrow). Scale Bar 20μM C (iii) Intracellular amastigote count for 4Hrs and 24Hrs p.i. Each dot represent count from 100 infected- KC. N=6. D Amastigote count from Giemsa-stained image of KC-infected with LD amastigotes, isolated from the spleen of 28 days LD-infected-mice. Each dot represent mean from 100 infected- KCs. N=6. *** signifies p value < 0.0001, ** signifies p value <0.001, n.s non-significant.

LD-S and LD-R quenches host membrane cholesterol during initial entry. A

(i) KCs were pre-incubated with NBD-cholesterol for 16Hrs. Subsequently, KCs were infected with RFP expressing LD-S or LD-R for 4Hrs or left uninfected and imaged. NBD-cholesterol was excited at 488nm, and LDs are presented with magenta pseudo color. White arrow indicates the NBD- cholesterol within LD while yellow arrow represents LD nucleus. Scale bar 5μm. A (ii) Co- localization analysis of LD (magenta), and NBD-Cholesterol (green) from A (i), marked in grey transparent box. B (i) Representative Flow cytometry-based analysis of NBD in LD-amastigotes isolated from LD-infected-KCs 2Hrs after infection pre-incubated with NBD cholesterol. Black Box representing % NBD positive population. B (ii) Graphical representation of MFI for three independent experiments as in B (i). n.s non-significant.

LD-R-PV have higher accumulation of Cholesterol and Fatty acids.

(A) Detection of PV Cholesterol, isolated from LD-S or LD-R-infected KC. Data represented as μg of protein/ μg of Cholesterol, (N=6). (B) TLC profile of isolated lipids from 24Hrs LD-infected-PVs with Cholesterol standard kept as control. (C) Individual Raman spectra from six independent LD- infected-KCs (i) LD-S-infected and (ii) LD-R-infected. Lipid-related peaks are demarcated within the spectra, including those at 1.540-560 cm-1 (Cholesterol), 2. 1080-1090 cm-1 (phospholipids), 3. 1270-1280 cm-1 (triglycerides), 4. 1300-1340 cm-1 (Amide-(II) I bond), 5. 1440-1453 cm-1 (Fatty acids and Triglycerides), and 6. 1650-1660 cm-1 (Amide-I bond). (D) Area under the curve, and shift are calculated and plotted for each experimental condition as represented in C (i) and (i). ** signifies p value <0.001,

Latrunculin-A restricts extracellular LDL influx inhibiting LD-R amastigote proliferation.

(A) Western blot comparing of LDLr expression between uninfected, LD-S, and LD- R-infected-KC 24Hrs p.i. B (i) Expression of LDLr determined by immunofluorescence in PECs treated with LDLr siRNA and scrambled siRNA (Sc). Scale bar 10μm. B (ii) Graph representing number of LD-amastigotes in LDLr knockdown per 100 PECs for B (i) (N = 6). Data: mean ± SD. C (i) Cofilin and phosphorylated-Cofilin expression by Western blot in LD-infected-PEC. C (ii) Graphical representation of comparative expression of abundance of Cofilin and phosphorylated- Cofilin with respect to β-actin for C(i). (D) Giemsa stained images of Latrunculin-A (LAT-A) and Cytochalasin-D (CYT-D) treated LD-infected-KCs in grey scale. LD nucleus are represented with Black arrow and host nucleus are represented as N. Scale bar 20μm. n.s non-significant.

Aspirin treatment inhibits Lipid droplets formation and increase responsiveness of LD-R amastigotes towards Amphotericin-B.

(A) Determination of lipid droplet accumulation by Nile red staining in LD-S and LD-R-infected-KC or KC left uninfected. Dotted line mark the cell periphery while dotted box cells were used for 3D reconstruction and represent in Movie 3. Scale Bar 5 μm. (B) Graphical representation of Lipid Droplet positive KCs in high lipid and low lipid media. Data represented as % of lipid droplet positive KCs (N=6) collected from 3 independent experiments. (C) Effect of Aspirin(5μM) on Lipid droplets accumulation in LD-R-infected-KCs. Independent infection was performed with LD-R1 and LD-R2 strains (N=6) and % KCs with high lipid droplets accumulation are represented. (D) Representative Giemsa stained images of LD-R- infected-KCs in different experimental conditions either untreated or treated with Aspirin (5μM) or Amp-B (0.36 μM) or with a combination of Aspirin (5μM) and Amp-B (0.11 μM) at 48Hrs p.i. Scale bar 20μm. *** signifies p value < 0.0001